Clustering Under Radius Constraints Using Minimum Dominating Sets

In this paper, we evaluate the applicability of algorithms designed to solve the minimum dominating set problem to perform clustering. The associated clustering problem relies on user constraints, and more specifically on radius intra-cluster constraints. We adapt and evaluate implementations from the state of the art on classification datasets, to compare them with other exact or approximate radius-based clustering algorithms, namely equiwide clustering and hierarchical agglomerative clustering with minimax linkage. We consequently provide the benchmark tools and datasets used in this work.